[VEP-0002] DataLink semantics vocabulary enhacement proposal = new DataDocumentation term

[Carlos Rodrigo]

(I reply only to the dal list)

> If Carlos wants to link "associated" publications, that's pretty
> certainly not debugging, except if it's technical publication on,
> say, a code or a spectrograph; but those probably should be confined
> to a reference URL or so in the first place and not be in datalink
> documents, right?

I don't think so (at least not in general)

And yes, Carlos wants, in some cases, to link associated publications to a given record :-)

My "use case" is more related to publications about how the record (spectrum, image, object distance
in a catalogue...) was generated. Just to mention, I think that we all know that the authors of the
works that have been necessary to produce those records usually like that their work is recognized
and that the corresponding references appear in a place as relevant as possible.

And, within the same service, the list of references associated to a particular record could be
different from the list of references associated to another record (for instance, in the case that I
have a service providing spectra measured by different authors). This doesn't always reduce to a
global "reference url".

And I think that something like DataLink, being a protocol to give a list of links associated to a
record, could include the possibility to give the links to the actual publications.


> So, let's assume this is about a publication derived from the
> observation in question: Where and how would a client display that?
> It's pretty clear to me that this doesn't belong to "debugging".
> It's more, well, "show me what happened later", as in progenitor.
> 
> So -- given these two use cases, perhaps we need
> 
> * generation-document (perhaps as a child of calibration or progenitor?)
>   and 
> * derivation-document (as a child of derivation), 

I guess that my use case is more related to the "generation-document" idea (though not necessarily
related to any calibration).

If people think that it is worth to make a difference between generation and derivation document, it
could be nice. But, if we try to follow your idea of keeping the list of terms as short as possible
unless necessary, I would just keep "documentation". And think about childs in the future if necessary.

Carlos